Tiny magnetic inductors delivered more than twice the inductance of comparable air-core devices—a gain that could help shrink chip power components, pending testing in complete converters. doi.org/hcq2bp
Tiny magnetic inductors delivered more than twice the inductance of comparable air-core devices—a gain that could help shrink chip power components, pending testing in complete converters. doi.org/hcq2bp
My first transistor radio, now long left behind, had 4 transistors in it. A top level iphone has more than 15 billion transistors. The music is a little better now.
October 27, 2026 at 8:23 PM
My first transistor radio, now long left behind, had 4 transistors in it. A top level iphone has more than 15 billion transistors. The music is a little better now.
It's the same as EEs using micro controllers and PLCs vs some relays or transistors for some simple interlock logic. It feels wasteful and less reliable for no apparent gain… but I guess that's the world we live in.
November 1, 2026 at 2:37 PM
It's the same as EEs using micro controllers and PLCs vs some relays or transistors for some simple interlock logic. It feels wasteful and less reliable for no apparent gain… but I guess that's the world we live in.
And we can't understand EVERYTHING, we're not gods, so we rely on abstraction. Computing is all voltages across transistors, but a proglang let's us not think about that and focus on modeling the logic, and we rely on the compiler to do that translation across the abstraction
November 3, 2026 at 2:49 PM
And we can't understand EVERYTHING, we're not gods, so we rely on abstraction. Computing is all voltages across transistors, but a proglang let's us not think about that and focus on modeling the logic, and we rely on the compiler to do that translation across the abstraction
Yes. But the difference is: Those three companies build the RAM chips with transistors several magnitudes smaller than what he did there in his shed. The problem isn't so much the conceptual capacity to do that, but the dimensions we need today.
Yes. But the difference is: Those three companies build the RAM chips with transistors several magnitudes smaller than what he did there in his shed. The problem isn't so much the conceptual capacity to do that, but the dimensions we need today.
Two Georgia teachers are bringing their Georgia Tech research experiences into the classroom, inspiring students to experiment, ask questions, and see new possibilities. #ResearchforRealLifeb.gatech.edu/3UyBhZ2
November 3, 2026 at 3:48 PM
Two Georgia teachers are bringing their Georgia Tech research experiences into the classroom, inspiring students to experiment, ask questions, and see new possibilities. #ResearchforRealLifeb.gatech.edu/3UyBhZ2
Radiation in LEO is 100s of times higher than on Earth. The chips used for AI are very, very sensitive to radiation because they use such tiny transistors. Everything that makes effective shielding is comically heavy, getting things into space remains very expensive.
November 2, 2026 at 10:21 PM
Radiation in LEO is 100s of times higher than on Earth. The chips used for AI are very, very sensitive to radiation because they use such tiny transistors. Everything that makes effective shielding is comically heavy, getting things into space remains very expensive.